They haven`t grabbed a lot of headlines, but data storage systems are getting hot
By Paul Demery
Shoppers today want flexibility in how, when and where they shop, and many
retailers oblige them with a growing range of online and offline options and
services. At retailers who have figured out how to execute these offerings well,
this has increased customer satisfaction levels and driven up sales.
It also has created a ton of customer buying data. And that, in turn, has
created a booming market for data storage technology—a market projected to grow
30% a year for the next five years. “We’re seeing a lot of storage growth, because
companies really want the agility to react quickly to changing business,” says
Phil Goodwin, a storage technology analyst with Meta Group Inc., Stamford, Conn.
As retailers learn how customers shop in different channels—and how they shop
across channels—the whole area of customer data is emerging as an invaluable
tool for perfecting multi-channel strategies. Customer relationship management
applications, highly segmented marketing campaigns, loyalty programs and customer
lifetime value calculations are only a few of the many different kinds of data
that retailers are collecting now that they weren’t a few years ago.
That massive build-up of data is also becoming a major challenge for retailers’
back-end storage systems, many of which were not designed to support the kind
of multi-channel shopping activity that many merchants are now experiencing.
Storage isn’t sexy, but without it, retailers lose out on a gold mine of information.
Data avalanche
Demand from retailers is part of the reason that research and analysis firm
Hurwitz Group, Framingham, Mass., projects sales of storage resource management
technology to grow a compound annual growth rate of 30% between 2001 and 2005,
from $397 million to $1.47 billion.
Much of the data captured by retailers relate to customer buying behavior,
showing streams of shopping preferences for individual customers. Retailers
mine these data within each selling channel for insight on how to better market
products within each selling channel. But to make the most use of the data,
retailers need a central storage environment where the data can be combined
with data from other channels, including catalog, web and stores.
“This data avalanche has to be centralized and consolidated,” says Liz Thibeault,
global manager, retail industry, EMC Corp., Hopkinton, Mass. “There is not a
retailer that doesn’t bring up customer data today, and particularly the amount
of data coming from online, where it accumulates much faster than offline.”
Noting that multi-channel shoppers generate substantially more revenue than
single-channel shoppers, retailers recognize a need “for any multi-channel retailer
to pull information from all channels,” she says.
The rise of the Internet is what’s behind the rise in data storage needs,
analysts say. The Internet has made it possible for retailers to gather extremely
detailed information about how their customers shop. And since that information
is important to understanding how to relate to customers, it can’t just be analyzed
and dumped. It’s all got be stored somewhere. “Web-enabled applications are
always hitting servers, and every time you do that, storage is involved,” says
Stephen Elliott, research director for storage management technology at Hurwitz
Group.
Trends in brick-and-mortar stores are also increasing the importance of storage
systems, as retailers move into such developments as self-checkout and self-service
kiosks used in several areas, including researching and purchasing products,
e-learning programs and job applications. These are trends that will continue,
causing data to grow at an increasing rate, says Brian Slaughter, manager of
retail industry storage systems for Dell Computer Corp., Austin, Texas, a major
provider of storage hardware systems as well as a big user of storage to supports
its own online retail operations. “Information will drive retailers’ business
more than ever before,” he says.
A tripling in capacity
The demands for storage are growing so rapidly that retailers are taking more
than just incremental leaps as they boost their storage capacity. Sears, Roebuck
and Co., for example, tripled its data storage capacity earlier this year to
140 terabytes from 45 terabytes. The increase allows Sears to better manage
data among several applications, enabling it to more tightly integrate data
on customer buying behavior with point-of-sale data and inventory data. “The
result is that we can better correlate various data points and improve our decision-making
about assortments, promotions, margins and inventory,” says Jonathan Rand, director
of merchandise planning and reporting at Hoffmann Estates, Ill.-based Sears.
Effective data storage systems require more than just larger capacity, however,
and retailers are beginning to recognize and demand greater interoperability
among their storage servers.
If retailers can organize data on an integrated infrastructure of storage
hardware and software, experts say, they can reduce the cost of managing information
while increasing the amount of data they can manage.
Sorting though the alphabet soup of options, though, can be daunting. For
starters there are SANs— storage area networks—that companies use for extensive
storage capacity. SANs, which can be networked via the web, tap the power of
multiple storage servers. For specifically storing fixed content, such as customer
receipts and medical prescriptions, there is content addressed storage, or CAS.
Then to make all that data work together, SANs and CAS applications can be combined
in automated network storage—or ANS—platforms.
Learning from history
The combination of these technologies in an ANS platform, such as from EMC
Corp., can play a key role in maintaining multi-channel retailing strategies.
For example, product pricing can be updated on one server and replicated to
multiple servers in real time, so that each channel can have instantly updated
information. The same advantage in real-time multiple server updates is also
used at retailers like Home Depot Inc. in employee training and support systems,
which can distribute updates in product specifications to salespeople on the
floor as well as to back-office managers.
In web-based retailing, these networked storage systems can be used to update
pricing and promotions based on changing customer behavior data. The TimeFinder
storage management system from EMC, for instance, enables a retail manager to
leverage data-mining analytics based on customer shopping behavior records to
make pricing and promotional changes on a copy of a consumer web site. Once
the changes are made, they’re instantly synchronized on the actual transactional
page. “The consumer doesn’t have to hit a refresh button,” says Thibeault.
At the Bombay Co. Inc., which designs and markets home furnishings on the
web and in 422 stores in the U.S. and foreign markets, a new centralized data
management strategy that integrates a network-attached storage server into a
SAN provides for a central data repository that will scale up to greater storage
capacity as Bombay increases business in all of its channels, says Chris Carroll,
director of infrastructure. Bombay, based in Fort Worth, Texas, is integrating
a network-attached storage server from Dell with a Dell EMC SAN.
Bombay, which installed its storage management system out of a necessity to
consolidate its growing amount of data, expects it to provide opportunities
to learn how to better use data for tracking trends and customer shopping behavior,
Carroll says. “The more history we collect, the more we’ll learn how to use
our storage technology to better use the data,” he says.
One of the additional ways Carroll expects to leverage the new system is to
create a product image library, so that Bombay can better manage the images
and make sure it’s presenting a consistent view of products across multiple
media, including web, print ads and in-store signage. Without that library,
he says, Bombay must take a more laborious route of checking how images appear
in each medium to assure their consistency.
Expected increases in multi-channel activity will only add to Bombay ‘s reliance
on management of back-end storage. “We see a lot of opportunity there to give
people faster and better access to information,” Carroll says, adding that Bombay’s
storage system will become more important as the company expands both its web
and store operations.
Bombay works with several data respositories, but by consolidating the data
onto one back-end storage system, it will make it easier for managers to access
the particular data they need. “So they can make decisions in a faster pace,”
Carroll says.
New era of integration
In a study of enterprise storage systems conducted earlier this year, Meta
Group Inc. found that many companies appear to be similar to Bombay in that
they need to learn how to best leverage storage technology. Among the study’s
findings are that most business organizations do not have enforced storage policies
or central storage administration.
“There is a relatively weak understanding of how much storage is being managed,
little confidence in current backup/recovery practices, and a general lack of
knowledge about emerging storage technologies,” the study says. In the meantime,
however, storage technology is evolving to make it more usable.
Storage technology and the product strategies of data storage vendors have
been evolving to make storage systems more interoperable among different brands
and, at least theoretically, more effective at helping merchants manage data
across multiple channels and data points.
Before the surge in online retailing in recent years, and the subsequent rise
in multi-channel retailing, there was little if any concern about making storage
servers interoperable. For one thing, data storage was cheap, so it usually
was not considered a financial burden to simply increase storage where it was
needed. And without a need to integrate data among different channels, retailers
could just build up their storage capacity in silos to support individual applications.
But while retail strategies and the demand they make on storage have changed,
many retailers’ storage systems have not. Many existing retail industry storage
systems were installed as long as 10 years ago, when storage servers were not
designed for interoperability. “Open standards are definitely ringing a bell
for people who have felt trapped for a long time,” says Slaughter of Dell.
The practical result of having a network of storage servers that interoperate,
he and others say, is that retailers can mine and analyze data from multiple
channels as well as from multiple applications, including customer relationship
management and supply chain management systems, to plan promotions and inventory
levels to support multiple channels, both separately and as an interrelated
group.
Long way to go
“The retail business is tougher than ever, with aggressive competition in
both price and assortment. Our challenge is to ensure that customers find the
merchandise and service they want in our stores, while eliminating what they
don’t want,” says Rand of Sears. He adds that he needs the right data storage
system to enable Sears to access and manipulate data in a way that supports
its merchandising efforts faster than the competition.
Still, some analysts say the move toward interoperability among different
brands of storage servers has a way to go in terms of universal reliability.
“No one today really supplies heterogeneous storage management tools,” says
Meta Group’s Goodwin. “We’re at a point where storage management is really in
its early stages, about the 3rd inning of a 9-inning game.”
Much of the data storage market is the turf of leading vendors Hewlett-Packard
Co., IBM Corp., EMC Corp. and Veritas Software Corp. But there are also 100
or more storage technology companies that have cropped up in recent years, though
many are expected to disappear in a market shakeup and consolidation, says Elliott
of Hurwitz Group.
Innovations
Storage management systems are usually customized for each user, so prices
vary widely. For an example of entry-level prices, however, EMC’s Clariion CX400
networked storage system starts at $62,000. The 2-gigabyte system can handle
cached data at 680 megabytes per second.
A few start-ups, meanwhile, are attracting the attention of analysts for what
appear to be innovative attempts to improve storage management. Two examples,
says Elliott, are Campbell, Calif.-based OuterBay Technologies and New York-based
NewView Technologies Inc. While OuterBay has a patent pending on improving storage
integration with PeopleSoft and Oracle databases, NewView is working on improved
integration in Windows-based file management systems.
Elliott and others note that extensive competition among storage vendors is
going a long way toward creating more open and flexible storage systems. “That’s
good for users,” Goodwin says. With the number of options increasing, retailers
should carefully invest in storage systems only where they can find an immediate
return, he cautions. “Think strategically and buy tactically,” he says.
paul@verticalwebmedia.com
